www.elsevier.com/locate/worlddev
World Development Vol. 29, No. 3, pp. 427±442, 2001 Ó 2001 Elsevier Science Ltd. All rights reserved Printed in Great Britain 0305-750X/01/$ - see front matter
PII: S0305-750X(00)00109-1
Rural Nonfarm Incomes in Nicaragua LEONARDO CORRAL Inter-American Development Bank, Washington, DC, USA and THOMAS REARDON * Michigan State University, East Lansing, USA Summary. Ð This paper examines nonfarm incomes of rural Nicaraguan households using a nationwide survey (LSMS) from 1998. The key ®ndings are as follows. (a) Rural nonfarm income (RNFI) constitutes 41% of rural household incomes. (b) RNFI is much more important than farm wage-labor income. (c) RNFI tends to be relatively concentrated geographically and socioeconomically, toward the rural areas of the Managua zone and the Rest-of-Paci®c zone, which are denser in infrastructure and population, and toward the upper income quartile of rural households. This concentration implies high entry barriers and capital requirements for rural nonfarm activity that the poor are simply not equipped to overcome. Equipping the rural poor through training and acquisition of diverse forms of capital to have a chance at the higher return nonfarm jobs would be a major step toward helping them to share the bene®ts of the rural nonfarm economy. (d) Selfemployment (small enterprise) income in manufactures is very minor, probably due to the ease of obtaining manufactured goods from urban industries and imports. Wage employment constitutes the bulk of RNFI (despite it receiving little attention in development programs and debate). (e) Three-quarters of RNFI is in the service sector, and only one-quarter is from manufactures; that can be contrasted with the emphasis on small manufactures enterprises in rural development programs and research. (f) Education, road access, as well as access to electricity and water were found to be important to nonfarm incomes. Ó 2001 Elsevier Science Ltd. All rights reserved. Key words Ð rural nonfarm income, Nicaragua, microenterprise
1. INTRODUCTION Nicaragua is the second poorest country (after Haiti) in Latin America and the Caribbean. The 1998 annual report of the Nicaraguan Central Bank states that 76% of the rural population are poor. Generation of rural nonfarm employment for the poor has emerged as an important rural development policy topic, given that the demand for farm labor by commercial farms is waning over time (Baumeister, 1999). This paper explores two questions: (a) what are patterns in rural nonfarm incomes across zones and household strata? (b) What are the determinants of individual participation in household incomes from these activities? Prior studies of rural nonfarm incomes and employment includeÐNitlap an-UCA. (1995); Renzi and Agurto (1996); Davis, Carletto, and Sil (1997), and Ruben, Rodriquez, and Cortez 427
(1999). Most of these studies were based on surveys that were incomplete geographically or in terms of variables on which observations were collected or analyzed. Our study is based on a multitopic survey that treats in more detail the rural incomes in Nicaragua than most prior surveys. Our data source is the second Living Standards Measurement Survey (LSMS), undertaken in 1998 by the Nicaraguan National Institute of Statistics and Census with the technical and ®nancial support of the World Bank, United Nations Development Program, the
* We
thank the World Bank, Klaus Deininger, and Eduardo Zegarra for provision of the data, and the anonymous reviewer and Eduardo Baumeister, Julio Berdegue, and Klaus Deininger for useful comments and suggestions, and the FAO Oce for Latin America and the Caribbean and the Inter-American Development Bank for funding.
428
WORLD DEVELOPMENT
Swedish International Development Agency, and the Social Emergency Investment Fund. We proceed as follows. Section 2 discusses data and sample characteristics. Section 3 discusses patterns in household income sources. Section 4 examines the determinants of individual participation in the nonfarm sector and of sources of rural household incomes. Section 5 concludes with policy implications. 2. DATA AND SAMPLE CHARACTERISTICS (a) Data The nationally representative sample comprised 4,209 householdsÐ2,270 urban and 1,939 rural, with 1,861 rural households after exclusion of outliers. (Seventy-eight households' observations were dropped as outliers. These included 45 households with negative total income and 33 households with total income greater than the mean plus three standard deviations, or 32,382 cordobas.) The sampling universe included all households in Nicaragua at the time of the 1995 Housing and Population Census. ``Rural'' was de®ned by the LSMS as population concentrations of less than 1,000 inhabitants per village. Income variables were de®ned as following. Farm income is net income, equal to the value of all output whether sold, given away, or home consumed, less the value of farm inputs (nonhousehold labor plus nonlabor inputs). Wage employment income from any sector is calculated as gross income. Self-employment income from commerce is calculated as net inventory movement hence gross sales less operational costs. Self-employment income in other nonfarm activities is calculated as net income. (b) Country, zones, and household characteristics The rural household income per capita is US$315 as calculated from our nationwide rural sample. (Compare that to the ocial GNP per capita ®gure of US$400.) Nicaragua is relatively thinly populated: in 1990, Nicaragua had 15 persons/square kilometer, vs 22 in all developing countries, and 52 in neighboring Guatemala. It is also relatively urbanized: the share of urban population in the total population was 54.4% in 1995, vs 38 in Guatemala and
40.7% in Honduras. But in only 33 of 145 municipalities is the rural population in the minority. Thus urbanization is relatively concentrated in the Managua department and the rest of the Paci®c region (Baumeister, 1999). Nicaraguan farm holdings are extremely concentrated. The survey data, which do not control for land quality, show that two-thirds of the farms are small (below ®ve manzanas of owned plus rented land, with 1 manzana equalling 0.7 ha), and they operate less than one-twentieth of the farmland nationally. Large farms (50 manzanas and above) constitute only one-tenth of all farms, but control three-quarters of farmland. Thirty-eight percent of Nicaraguan rural households are landless, based on the survey data. Table 1 shows rural household characteristics derived from the survey data. The sample is strati®ed into four zones, including one zone de®ned administratively, the Managua zone (a department in the Paci®c region), and three zones de®ned agroclimatically (as the three ``macroregions'' in Baumeister, 1999)Ðthe Rest-of-Paci®c (the Paci®c Macroregion less the Managua Department), the Interior zone, and the Atlantic zone. Note that we set the Managua Department apart from the rest of the Paci®c macroregion because it contains the capital city and is substantially more densely populated and better served with infrastructure. The Rest-of-Paci®c zone comprises two agroecological zones. The western subzone (Le on and Chinandega departments) is more hot and dry than the eastern zone, and has volcanic soils. The western subzone has historically had substantial agro-export activity, mainly of sugar, peanuts, bananas, and soy. The eastern subzone has a cooler climate and produces coee, pineapple, and vegetables. The Rest-of-Paci®c zone has fertile soils that permit intensive production of annual crops. This zone also has the best infrastructure and rural markets. The Atlantic zone covers more than 30% of Nicaragua, and is mostly humid tropical forests and mangrove swamps. There are several extensive ranching and subsistence agriculture areas. Infrastructure is poor, and markets for products and land are little developed. The Interior zone (which includes the Northern and Central subzones) has the most mountainous topography, and thus the greatest erosion problems. Its infrastructure is in third rank, after the Managua and Rest-of-Paci®c zones. Its agriculture includes coee, vegetables, tobacco, and livestock.
NICARAGUA
429
Table 1. Characteristics of rural households by Zone, 1998 All
Managua Zonea
Rest-ofPaci®cb
Interiorc
Atlanticd
Number of households (%)
1861
3.2
33.1
44.1
19.6
Land holdings Avg. farm size (manzanas) Landlessness rate (% of households) % of households with 0>5 mzs
14.1 37.9 33.4
6.8 58.3 26.7
6.1 52.2 31.2
9.2 35.1 39.0
39.8 16.7 26.0
82.2 2.8 0.4 44.9 3.7 61.2 2.9
78.3 0.0 0.0 47.3 3.7 77.5 4.9
80.5 0.0 0.0 46.5 3.8 71.8 3.8
81.3 0.0 0.0 44.5 3.7 56.9 2.5
86.8 14.5 2.2 42.7 3.6 51.3 2.1
8.9 33.2 4.9 2.7
1.7 46.7 16.7 11.7
11.9 44.9 9.3 4.7
8.7 28.7 2.6 1.8
4.1 21.9 2.2 0.8
8.1 2.1
8.3 1.7
9.3 4.4
8.9 1.2
4.4 0.3
Infrastructure access (% households) Access to household Paved road Dirt road Trail Electricity hook-up Water hook-up
7.2 52.8 28.8 29.5 19.6
6.7 90.0 0.0 83.3 56.7
9.8 70.6 14.1 51.6 36.7
8.0 43.0 36.7 21.9 12.4
0.3 41.4 38.1 5.8 4.1
Social and organizational capitalg Community related social capital Business related social capital
30.5 2.1
51.7 8.3
28.7 2.2
33.0 1.8
28.2 2.5
Managua vs Paci®c
Managua vs Interior
Managua vs Atlantic
Paci®c vs Interior
nsh ns ns
ns ns ns
ns
ns
ns ns ns
ns ns ns
ns
Labor and demographic Gender of household head (% men) Household head Miskito (%) Household head Creole (%) Age of household head Number of adults (>12 years) Adults who can read and write (%) Ave. years of education of adultse Degree of most educated member Preschool (% of households) Primary (% of households) Secondary (% of households) University (% of households) Incidence of migration tof Domestic (%) Foreign (%)
T tests on the equality of means for Table 1
Land holdings Average farm size (manzanas) Owned Rented Labor and demographic Gender of household head (% men) Age of household head Number of adults (>12 years) Average years of education of adults Adults who can read and write (%) Degree of most educated member Preschool (% of households) Primary (% of households) Secondary (% of households) University (% of households) Incidence of migration to Domestic (%) Foreign (%)
ns
Paci®c Interior vs Atlantic vs Atlantic
ns
ns
ns
ns
ns ns
ns ns
ns ns
ns ns
ns
ns
ns Continued next page
430
WORLD DEVELOPMENT
Table 1Ðcontinued Managua vs Paci®c Infrastructure access (% households) Access to household Paved road Dirt road Trail Electricity hook-up Water hook-up
Managua vs Interior
Managua vs Atlantic
Paci®c Interior vs Atlantic vs Atlantic
ns ns
ns
ns ns
ns
ns
Social and organizational capital (% of households) Community related social capital Business related social capital
ns
Paci®c vs Interior
ns
a
The Managua zone includes the department of Managua. The ``Rest-of-Paci®c'' includes the departments Chinandega, Le on, Masaya, Granada, Carazo, and Rivas. c The ``Interior'' includes the departments Nueva Segovia, Jinotega, Madriz, Esteli, Matagalpa, Boaco, and Chontales. d The Atlantic zone includes the departments Rõo San Juan, RAAN and RAAS. e Imputed years of education for each member were obtained in the following manner: if highest degree was: Preschool and knows how to read and write: 3 years; Preschool and does not read or write: 1 year; Primary: 6 years; Secondary: 11 years; Basic Technical school: 6 years; Middle Technical school: 9 years; Superior Technical school: 12 years; University: 16 years. To these imputed years the approved years in their current studies, if any, were added. Household average years of education of adults were obtained by averaging over household members who are 12 years or older. f Incidence of migration represents a household member reporting having changed residential location for work reasons over the past 12 months. g Social and organizational capital indicates whether a household member participated in a community organization or group, such as religious organizations, parent associations, neighborhood committees or municipal development committee, or if a household member participates in a professional association or a loan and savings cooperative. h nsÐnot statistically signi®cantly dierent from each other. * Area statistically dierent from each other at 10% level. ** Area statistically dierent from each other at 5% level. *** Area statistically dierent from each other at 1% level. b
As one moves from the rural areas of Managua and the Rest-of-Paci®c zones toward the Interior and then Atlantic zones, hence from near to the country's capital and ports to furthest from them, poverty increases. According to a survey by the Ministry of Social Action in 1997 (MAG-FOR, 1999), 64% of the population is poor in the Atlantic Zone and 72% are poor in the Interior. As shown in Table 1: (i) average farm size increases, from 6 mz in the Rest-of-Paci®c, to 9.2 in the Interior, and 39.8 mz in the Atlantic zone; (ii) infrastructure (road, electricity, and potable water) and population density decline; (iii) the population share of the landless declines (from half to about a third); (iv) literacy and education decrease; (v) business- and community-related social capital (such as associations) declines; (vi) migration rates decrease although overall they are lower than might be expected. Note that only 8% of households have migrants to other parts of Nicaragua and only 2% have
migrants outside Nicaragua for the survey year. One-quarter of those undertaking migratory employment do so as farm workers, and threequarters as nonfarm workers. There is a relative geographic concentration of foreign migration among households in the Managua and Restof-Paci®c zones. Moreover, the poor tend to migrate to farm jobs as these require the least quali®cations but oer the lowest pay. 3. PATTERNS IN NONFARM EARNINGS (a) Nationwide patterns Table 2 shows national patterns of income sources of rural households by land and income strata. Total sample ®gures are averages over the strata, weighting by strata populations. The average share of nonfarm income in total earned income (excluding transfers and pensions) is 41%. Despite the importance for
Table 2. Income/capita sources of rural Nicaraguan households by land and income strata Own-farm income
Agriculture wage income
Nonfarm wage income
Nonfarm selfemployment
Rental income
Pensions
Other income
Total income
Total income 1998 Cordobas
All
1861
Land strata 0 mzs
35 (2.2)a
17 (2.4)
30 (2.5)
11 (4.1)
0 (19.2)
0 (15.3)
7 (4.0)
100 100
3450 (1.2)
37.9 12.7
2 < 5 mzs
20.7
5 < 20 mzs
14.6
20 < 50 mzs
6.4
>50 mzs
7.6
23 (2.0) 24 (1.8) 18 (3.2) 9 (2.6) 5 (2.5) 5 (3.6)
52 (1.8) 21 (2.4) 17 (3.0) 11 (3.2) 15 (3.1) 3 (3.1)
13 (3.5) 8 (4.0) 7 (5.0) 9 (6.0) 15 (3.1) 7 (3.4)
0 (15.9) 0 (14.9) 0 ) 0 (12.8) 0 (11.0) 0 )
0 (10.9) 0 (12.6) 0 (19.2) 0 (17.0) 0 ) 0 (11.7)
8 (3.3) 6 (5.7) 6 (3.5) 6 (3.1) 9 (4.1) 7 (6.1)
100
